THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 264. No , 20, Issue of ‘July 15, pp. 11671-11681.1989 Printed in U. S. A .

Interaction of the y-Subunit of Retinal Rod Outer Segment Phosphodiesterase with Transducin USE OF SYNTHETIC PEPTIDES AS FUNCTIONAL PROBES*

(Received for publication, March 1, 1989)

Daniel F. Morrison, Jess M. Cunnick, Brenda Oppert, and Dolores J. Takernoto$ From the Department of Biochemistry, Kansas State University, Manhattan, Kansas 66506

There is considerable evidence which suggests that the y-subunit of cGMP phosphodiesterase (PDEy) is a multifunctional protein which may interact directly with both the catalytic subunits of PDE (PDEaB) and the a-subunit of transducin (Ta) (Whalen, M., and Bi- tensky, M. (1989) Biochem. J. 259, 13-19; Griswold- Prenner, I., Young, J. H., Yamane, H. K., and Fung, B. K.-K. (1988) Invest. Ophthalmol. & Visual Sei. 29, (Suppl.) 218). To determine the region of interaction between the multifunctional PDEy and Ta, and to de- termine the significance of this interaction, peptides corresponding to various regions of PDEy were syn- thesized and tested for their ability to inhibit the GTPase activity of Ta. One of these peptides, PDEy-3 (bovine amino acid residues 31-45), inhibited the GTPase activity of Ta with an Iao of 450 MM. The peptide (PDEy-3) was found to inhibit the GTPase activity of Ta by inducing the binding of transducin to the rod outer segment membrane and by altering the GTP/GDP exchange. Analogs of PDEy-3 were synthe- sized to determine the required structure of the PDEy- 3 region needed for the interaction of PDEy with Ta. The results of these studies indicated that the removal of the positively charged amino acids or any of the potential hydrogen-bonding amino acids increased the Iao for the inhibition of the GTPase activity of Ta. Substitution of the hydrophobic amino acids had no effect. These results indicate the hydrophilic interac- tions may be essential for the binding of PDEy to Ta and for the inhibition of the GTPase activity of Ta by PDEy. The observed effects of PDEy-3 on Ta and on PDE suggest that PDEy is a multifunctional protein which may play more than one role in the deactivation of the retinal transduction cascade.

Retinal rod outer segments contain a light-activated guanosine 3’,5’-cyclic monophosphate phosphodiesterase

* This project was supported in part by Grant EY06490 from the National Institutes of Health/NEI (to D. J. T.) and in part by Grant BRSG SO7 RR 05373 from the Biomedical Research Grant Program, Division of Research Resources, National Institutes of Health. This is Contribution 89-348-5 from the Kansas Agricultural Experiment Station. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$To whom correspondence should be addressed. Tel.: 913-532- 7009.

(PDE)’ which hydrolyzes cGMP in the rod outer segment (1- 3). Accompanied by the decrease in cGMP is a decrease in sodium conductance of the rod outer segment plasma mem- brane (4). The decrease in the sodium flux through the rod outer segment plasma membrane has been directly connected to the level of cGMP in the rod outer segment (4-6).

It was formerly thought that the bovine rod outer segment PDE complex was made up of three subunits with the struc- ture PDEaPy (7). However, experiments performed by De- terre et al. (12) have suggested that the PDE complex may have the structure of PDEa@y2. The molecular weights for a, p, and y derived from sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) are 88,000, 84,000, and 11,000, respectively (7). The molecular weight of 9,700 for the y-subunit of PDE (PDEy), derived from sequence analysis (9), agrees reasonably well with the molecular weight deter- mined by SDS-PAGE (7).

Evidence indicates that PDEy inhibits PDE catalytic activ- ity (8). Partial trypsin digestion of PDEaPyz leads to a deg- radation of PDEy, paralleled by an increase in PDE catalytic activity (8). The degree of increase in PDE catalytic activity is directly proportional to the amount of PDEy degraded (8). Readdition of purified PDEy to this trypsin-activated PDE results in a 97% inhibition of PDE catalytic activity (8).

The full activation of PDE requires the removal of both PDEy subunits (12). When PDE is activated, the a-subunit of transducin (Ta) interacts with and possibly displaces PDEy from the a- and p-subunits of PDE (PDEaP) (10). Evidence suggests that Ta may bind directly to PDEy to facilitate its removal from PDEaP (12, 13). Ta may also bind to PDEaP as well (11). The removal of one PDEy subunit from PDEaP requires stoichiometric amounts of Ta and results in only a partial activation of the catalytic activity of PDEaP (12). Removal of the second PDEy subunit is depend- ent upon the concentration of Ta ( i e . more than 100 Ta per PDEapy complex is needed for the removal of the second PDEy) and fully activates the catalytic activity of PDEaP (12). This suggests that the affinities of the two PDEy sub- units for PDEaP may be different (12).

The amino acid sequence of the bovine PDEy subunit has

The abbreviations used are: PDE, retinal rod outer segment guanosine-3’,5’-cyclic monophosphate phosphodiesterase; PDEa, PDEP, PDEy, the three subunits of the PDE; GMPP(NH)P, guanyl- 5”yl imidodiphosphate; GTPyS, guanosine 5’-O-(thiotriphosphate); 8-ME, 2-mercaptoethanol; PMSF, phenylmethylsulfonyl fluoride; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electropho- resis; RIA, radioimmunoassay; IRBP, interstitial retinol binding pro- tein; BSA, bovine serum albumin; MOPS, 3-(N-morpholino) propanesulfonic acid; ROS, rod outer segment; HPLC, high perform- ance liquid chromatography.

11671

11672 Effect of Retinal PDE-y on Transducin

been determined by protein and cDNA sequencing (9). We have previously reported using this sequence to synthesize peptides corresponding to various regions of the entire PDEy subunit (18). These peptides were tested for their ability to inhibit both PDE catalytic activity and the GTPase activity of Ta. The results of these experiments indicated that the peptide PDEy-3 (bovine amino acid residues 31-45) was the most effective peptide for inhibition of the GTPase activity of Ta. In this report, further data were presented which detail a possible mechanism by which the PDEr subunit inhibits the GTPase activity of To.

EXPERIMENTAL PROCEDURES

Materials-Fresh bovine eyes were obtained from a local slaugh- terhouse (Iowa Beef Packers. Emnoria. KS). Buffers. euanosine 5’- triphosphate (GTP), 5’guanylyhmidodiphosphate (GMPP(NH)P), guanosine 5’-diphosphate (GDP), and other reagents were from Sigma. Guanosine-5’-0-(thiotriphosphate) (GTPrS) was from Boeh- ringer-Mannheim. [y-3ZP]GTP (50 Ci/mmol), [3SS]GTP$S (1200 Ci/ mmol), and [8-3H]GDP (9 Ci/mmol) were from Du Pont-New Eng- land Nuclear. [8-3H]GMPP(NH)P (5 Ci/mmol) and [8-aH]cGMP (19 Ci/mmol) were from ICN Radiochemicals. Carrier-free “‘1 was from Amersham Corp. The t-butoxycarbonyl amino acids and their resins were from Vega Biochemicals, United States Biochemicals Corp., Cleveland, OH, or Sigma. Aquacide III was from Behring Diagnostics. All other reagents used for peptide synthesis were HPLC grade from Fisher or Sigma, or Sequenal grade-from Pierce. All materials used for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) were from Bio-Rad or Sigma.

Preparation of UndepZeted Rod Outer Segments-Rod outer seg- ments were prepared by the method of Papermaster and Dreyer (21). Fresh bovine eyes were obtained from a local slaughterhouse within 60 min of slaughter. The eyes were transported in the dark and on ice. Retinas were removed under dim red light and stored without buffer at 70 “C.

Typically, a preparation consisted of 50-100 dark-adapted retinas thawed over several hours in the dark. All purification steps were done in dim red light and on ice unless otherwise indicated. The retinas were suspended in ROS-No. 1 (5 mM Tris (pH 7.4), 2 mM MgCl*, 65 mM NaCl, 1 mM 2-mercaptoethanol (P-ME), 0.5 mM phenylmethylsulfonyl fluoride (PMSF), and 34% sucrose (w/w (d = 1.15)) using 1 ml of buffer per retina. The retinas were shaken vigorously for 1 min, shearing the rod outer segments from the rod inner segments (21). The rod outer segments were isolated from the total retina by centrifugation at 4000 rpm for 4 min in a SS-34 rotor (Sorvall). The resulting pellet was resuspended in ROS-No. l(15 ml/ tube) and recentrifuged as above.

The suspended rod outer segments were combined and diluted to three times their volume in RCS-No. 2 (10 mM Tris (pH 7.4), 1 mM M&h. 1 mM B-ME. and 0.5 mM PMSF). The rod outer sezments

m - I I

were pelleted at 10,000 rpm for 10 min in a SS-34 rotor, then resuspended in 10 ml of ROS-No. 3 (10 mM Tris (pH 7.4), 1 mM MgC12, 1 mM S-ME, 0.5 mM PMSF, and 26.3% sucrose (w/w).

The suspended rod outer segments were layered (3 ml) on a discontinuous sucrose density gradient of 1 ml each of 1.11,1.13, and 1.15 g/ml of sucrose in ROS-No. 2. The preparation was centrifuged at 24.000 rum for 45 min in a SW-27.1 rotor (Beckman).

Purified-rod outer segments were removed from the gradient at the 1.11-1.13 g/ml interface and washed with ROS-No. 7 (10 mM Tris (pH 7.4), 100 mM NaCl, 5 mM MgCl*, 0.1 mM &ME, and 1 mM PMSF). The rod outer segments were pelleted at 14,000 rpm for 15 min in a SS-34 rotor. This procedure was repeated 2 more times. The washed pellet was resuspended in GTPase buffer (10 mM 3-(4-mor- pholino)propanesulfonic acid (MOPS) (pH 7.4), 2 mM MgCl,, 1 mM P-ME, and 0.1 mM PMSF) for subsequent assays. The rod outer segment preparation was stored at -70 “C and is referred to, herein, as an undepleted rod outer segment preparation.

To make PDE-depleted rod outer segment membranes, the washed pellet obtained from the above procedure was resuspended in 50 ml of PDE extraction buffer (5 mM Tris (pH 7.4), 0.5 mM MgC12, 1 mM S-ME. and 0.1 mM PMSF) (24). The rod outer segment membranes were ruptured by passing the suspension through; 22-gauge needle three times. The suspended rod outer segment membranes were exposed to room light, on ice for 30 min, then pelleted at 16,000 rpm for 30 min in a SS-34 rotor. This procedure was repeated twice.

GTPase Assay-The GTPase assay was performed as previously described (22, 23). Typically, 4 pg of undepleted rod outer segments were used per 100 ~1 of GTPase buffer. Reactions were initiated by the addition of GTP to a final concentration of 20 NM GTP (20 pM GTP + 0.1 pCi of [Y-~‘P]GTP per assay tube). The reaction mixture was incubated under room light at 37 “C for 20 min, then terminated by the addition of 1 ml of a solution containing 150 mM perchloric acid, 0.1 mM KH2P0,, and 1 ml of a solution containing 10 mM ammonium molybdate, 20 mM triethylamine-HC1. The resulting pre- cipitate was collected on a Whatman GF/A glass microfiber filter using a suction filtration apparatus. Each filter was washed with 50 ml of a solution containing 10 mM triethylamine-HCl, 200 mM perchloric acid, 2.5 mM ammonium molybdate. The quantity of “‘Pi released from the y-phosphate of GTP was determined by liquid scintillation counting. In the kinetic experiments, the GTP concen- tration was varied from 10 to 100 pM (final concentration). In some experiments, peptides were preincubated with the undepleted rod outer segment membranes for 5 min at 37 “C prior to the addition of substrate.

PDE Assay-PDE activity was determined by the method of Thompson and Appleman (50). The final concentration of the reac- tion was 50 mM Tris (pH 7.4), 5 mM MgC12,40 pM cGMP, [3H]cGMP (150,000 cpm/tube, specific activity 19 Ci/mmol) in a final volume of 400 pl. Reactions were at 30 “C. Purified PDEy was added as indi- cated.

Preparation of PDE-y-PDEr was purified as previously described (51). Crude PDE was eluted under room light from undepleted rod outer segments in 10 mM Tris (pH 7.4), 0.1~mM dithioerythritol, 0.5 mM MgCl,, 1 pM leupeptin, 0.2 mM PMSF, 1 pM pepstatin. The eluate was concentrated with Aquacide III and the PDEr was pre- pared as described (51).

Prelabeling TCX with p?S]GTPyS-The a-subunit of transducin can be eluted from rod outer segment membranes and labeled with nonhydrolyzable analogs of GTP (24). Rather than using [8-3H] GMPP(NH)P (24), another nonhydrolyzable analog of GTP [a’SI GTP+l, was used. This analog was used so that subsequent dual labeling would be possible. PDE-depleted rod outer segment mem- branes (25 mg) were diluted 1:l (final volume = 2 ml) with GTPase buffer which contained 600 mM NaCl and 300 pM GTP-yS (300 pM GTPrS + 1250 ltCi of 135S1GTP~S) and incubated on ice under room light for 12 h. The ROSmembranes.were then pelleted by centrifuging at 16,000 rpm for 30 min in an SS-34 rotor. The supernatant, containing [35S]GTP-yS-labeled Trr (Tel-[?S]GTPrS), was analyzed by SDS-PAGE to check for purity. Transducin (10 mg/ml) was found to be greater than 85% pure when stained with Coomassie Blue. The partially purified Tm-l?S]GTPrS was then used in the nucleotide release assay described below.

Nucleotide Release Assay-Ten pl (10 pg) of the partially purified Ta-[?S]GTPrS was incubated in the presence of various quantities of the neotide. PDEr-3. in 100 ul of GTPase buffer for 10 min. at 37 ‘C. Samples were’filtered through a Millipore HAWP filter pre- washed with GTPase buffer. The filter was then washed three times with 1 ml of GTPase buffer, and bound nucleotide was quantitated by liquid scintillation counting.

[8-3H]GMPP(NH)P Binding Assay-[8-3H]GMPP(NH)P binding was assaved bv the method of Stayer et al. (24). An assay consisted of 50 pg”of &depleted rod outer segment membranes in 100 ~1 of GMPP(NH)P-binding buffer (10 mM MOPS (pH 7.4), 60 mM KCl, 30 mM NaCl, 2 mM MgC12, 1 mM p-ME, 0.1 mM PMSF, and 1 mg/ ml BSA). The reaction was initiated by the addition of GMPP(NH)P to a final concentration of 20 uM (20 UM GMPP(NH)P + 0.1 &i of [8-3H]GMPP(NH)P per assay tube). The reaction mixture was in- cubated at room temperature under room light for 30 min. After 30 min, the reaction mixture was filtered through a 0.45~pm Millipore HAWP filter, using a suction filtration apparatus, and washed three times with 1 ml of GMPP(NH)P-binding buffer. The quantity of [8- 3H]GMPP(NH)P bound to the filter was determined by liquid scin- tillation counting. In some experiments, undepleted rod outer seg- ment membranes were preincubated with peptides for 5 min at room temperature prior to the addition of substrate.

~?3/GTP~S/~H]GDP Binding Assay-A typical assay consisted of 60 pg of dark adapted undepleted rod outer segment membranes in 100 ~1 of GMPP(NH)P-binding buffer. [8-3H]GDP, [3SS]GTPyS, or both nucleotides (1 pM final concentration each - 1 pM each unlabeled nucleotide + 0.1 &i of each labeled nucleotide per assay tube) were added and the reaction initiated by illumination of the undepleted rod outer segment membranes under room light. The reaction mixture was incubated at 37 “C for 5 min under room light.

Effect of Retinal PDEr on Transducin 11673

The reaction was terminated by applying 85 p1 of the reaction mixture to a 0 . 4 5 - p ~ Millipore HAWP filter using a suction filtration appa- ratus. The filter was washed three times with GMPP(NH)P-binding buffer. The quantity of each labeled nucleotide bound to the filter was determined by liquid scintillation counting using a dual label program. In some experiments, undepleted rod outer segment mem- branes were preincubated with 200 nmol (2 mM) of the peptide, PDEy-3 at 37 "C in the dark for 5 min prior to the addition of substrate.

Membrane Reconstitution Assay-Transducin was eluted from PDE-depleted rod outer segment membranes (22, 26) with 5 ml of ROS-No. 8 (10 mM Tris (pH 7.5), 0.1 mM @-ME, 1 mM PMSF) containing 40 p~ GTP. The rod outer segment membranes were resuspended by passing them through a 22-gauge needle three times. The rod outer segment membranes were incubated on ice for 20 min and were then pelleted by centrifugation for 30 min at 16,000 rpm in a SS-34 rotor. This procedure was repeated three more times, and the resulting supernatants were combined to make a crude transducin preparation (about 85% pure). The rod outer segment membranes were washed six more times in the same manner as described above and the supernatants discarded. Both the crude transducin prepara- tion and this rod outer segment membrane preparation were tested for reactivity to various antipeptide antisera using the solid-phase radioimmunoassay (RIA) described below.

In the reconstitution assay, 17.5 pg of the depleted rod outer segment membrane preparation and 8 pg of the crude transducin preparation were incubated with and without 51 nmol (510 pM) of the peptide, PDEy-3, and 40 p~ GTP on a shaker at room tempera- ture for 10 min in the light (total volume = 100 p l ) . The rod outer segment membranes were pelleted by centrifugation in a microfuge (Fisher) for 2 min. The supernatants were removed and assayed using antipeptide antisera in a solid-phase RIA. Experiments included incubating 17.5 pg of the depleted rod outer segment preparation alone, and with and without 51 nmol of PDEy-3, and incubating 8 pg of the crude transducin preparation alone, and with and without 51 nmol of PDEy-3. Equivalent quantities of human lens membranes were used in place of the depleted rod outer segment membranes as negative controls for background subtraction in the solid-phase RIA. Lens membranes were prepared as described using the NaOH extrac- tion method (27).

Solid-phase Radioimmunoassay-The solid-phase RIA was a mod- ification of the method of Suter (28). Polystyrene tubes (12 X 75 mm) were filled with 0.5 ml of 0.2% glutaraldehyde in 0.1 M sodium phosphate, pH 5.0, and shaken for 3 h at room temperature. The tubes were then washed three times with 0.1 M sodium phosphate (pH 5.0). The supernatants from the above described experiments were diluted to 200 pl with 0.1 M sodium phosphate (pH 8.0), added to each tube, and incubated at 37 "C for 2 h. The tubes were then washed three times with 0.15 M NaCl, 0.5% Tween 20, to block nonspecific binding, and one time with water. Five hundred microli- ters of the desired antipeptide antisera diluted to 1:lOO in 0.1 M sodium phosphate, 0.9% NaCI, 0.05% Tween 20, 0.02% NaN3 (pH 7.2), was added and the tubes were shaken for 12 h at room temper- ature. The tubes were then washed three times with 10 mM Tris, 0.05% Tween 20 (pH 8.0), and two times with water. Five hundred microIiters of 0.1 M sodium phosphate, 0.9% NaCI, 0.05% Tween 20, 0.02% NaN3 (pH 7.2), containing 1 p1 of '251-labeled protein A/ml of buffer (1 X lo6 cpm/tube), was added and the tubes were shaken for 1 h at room temperature. The tubes were washed two times with 10 mM Tris, 0.05% Tween 20 (pH 8.0), two times with water, and then counted in a y counter (Beckman).

Peptide Synthesis and Purification-Peptides corresponding to 15- amino acid-long segments of bovine PDEy were synthesized manually by the method of Hodges and Merrifield (14) as modified by Gorman (15) except that cleavage of the peptide from the resin and protecting groups was accomplished using HBr in anhydrous trifluoroacetic acid (16) instead of anhydrous HF. HF cleavage was used for peptides which were analyzed for circular dichroism.

Peptides used in inhibition assays were purified by cellulose thin- layer electrophoresis in a buffer containing acetic acid/formic acid/ water (31:16). Peptides were spotted onto the thin layer cellulose plates and dried. Buffer was then sprayed onto the thin layer cellulose plates and electrophoresed in the same buffer a t 1000 V until the tracking dye, pyronin y, had migrated 9 cm. After electrophoresis, the plate was dried and the peptide was visualized with ninhydrin. Cor- responding unsprayed lanes were scraped, and the peptide was eluted with 0.5% NH4OH. The peptides were then lyophilized and dissolved

in distilled water. The pH of the peptides were adjusted to 7.4 with 2 M Tris (pH 11).

Peptides needing further purification for circular dichroism meas- urements were purified using reverse-phase HPLC on a Vydac C-18 column using a 0-100% 0.1% trifluoroacetic acid/acetonitrile gradient in 0.1% trifluoroacetic acid/distilled water. Peaks were detected by monitoring the ultraviolet (UV) absorption of the peptide at 230 nm. Peaks were collected and rerun to determine the purity of the collected peaks by monitoring the UV absorption of the peptide at 215 nm.

Peptides were quantitated by reverse-phase HPLC using a Vydak C-18 column and a 10-50% gradient of acetonitrile in 0.01 M sodium phosphate (pH 7.0) using o-phthalaldehyde as a detecting reagent (17). Peptides (5-pl aliquots) were hydrolyzed for 12 h in vacuo using 6 N HCl at 110 "C. The HCl was lyophilized off, and the amino acids were dissolved in 100 pl of water. Aliquots of 10 p1 were analyzed for amino acid content and were quantitated by comparison with known amino acid standards using a Shimadzu peak integrator.

Radioiodination of PDEy-3"An analog of PDEy-3, PDEy-3+Tyr, was utilized for iodination of the chloramine-T method (48). PDEy- 3+Tyr contains a tyrosine on the N terminus of the peptide. An aliquot of 30 pl of PDEy-3+Tyr (7.5 pg/ml) was diluted to 50 p1 with water. Fifty p1 of a 1 mM (NHJHCO, and 50 pl of a 1 mg/ml chloramine-T solution were added to a final volume of 150 pl. One pCi of carrier-free "'I was added and the reaction mixture incubated for 1 min. The reaction was terminated by the addition of a 1 mg/ml solution of sodium metabisulfite. The peptide was purified by two- dimensional thin layer chromatography (TLC). Iodinated peptide was spotted on the corner of a TLC plate and chromatographed with a buffer containing I-butanol/pyridine/acetic acid/water (1:10:2:8) in the first dimension, then, chromatographed with a buffer containing water/acetic acid/formic acid (16:3:1) in the second dimension. Ra- dioactive peptide was detected by autoradiography using Kodak XR- 1 film and DuPont Cronex screens. The peptide was scraped from the TLC plate, eluted with distilled water, and analyzed as described for unlabeled peptides.

Peptide Binding Assay-Various quantities of rod outer segment membranes, lens membranes, or BSA were incubated with 1.2 nmol of lZ5I-PDEy-3+Tyr for 10 min at 37 "C in 100 pl of GTPase buffer containing 1 mg/ml BSA. After 10 min, the mixture was filtered on a 0.45-pm Millipore HAWP filter prewashed with GTPase buffer containing 1 mg/ml BSA. The filters were washed with 1 ml of GTPase buffer containing 1 mg/ml BSA and then with 1 ml, and then 10 ml of GTPase buffer. The quantity of peptide bound to the filter was determined by counting in a y counter (Beckman).

For competition of '251-PDEr-3+Tyr binding by unlabeled pep- tides, 35 pg of rod outer segment membranes were preincubated for 5 min at 37 "C with unlabeled peptides. Then, 1.2 nmol of lZ6I-PDEy- 3fTyr was added and the mixture (total volume = 100 p1) incubated for 10 min at 37 "C. The reaction mixture was filtered, washed, and counted in the same manner as described above.

Circular Dichroism Measurement of PDEy-3 and PDEy-3 Ana- logs-Purified peptides were dissolved in 0.01 M sodium phosphate, 2 mM MgC12, pH 7.0, to a concentration of 0.5 mg/ml. Samples were scanned from 240 to 190 nm at a rate of 10 nm/min with a time constant of 4, at room temperature on a Jasco J-500A spectropolar- imeter using a 0.5-mm quartz cell. The circular dichroism spectrum of PDEy-3 was compared to standard spectra recorded by Greenfield and Fasman (39).

Miscellaneous Methods-Protein concentrations were determined by the method of Bradford (29) using BSA as a standard. SDS-PAGE was performed as described by Laemmli (25). Statistical p values were determined using Student's t test (49). The concentrations of individual proteins in crude rod outer segment preparations was determined by scanning of autoradiograms of preparations from Western blots. Antisera directed against each individual protein was reacted with the blot and visualized with '"I-protein A. Quantitation was accomplished by comparison with known concentrations of pu- rified proteins (such as purified Tor) using a Gilford multimedia densitometer and Shimadzu integrator.

RESULTS

Inhibition of the GTPase Activity of TLY by PDEr Protein- PDEy has been reported to inhibit PDE activity when added back to trypsin-treated PDE (8).

Fig. L4 depicts the inhibition of trypsin-activated PDE by exogenous PDEy. The 150 for PDEy was around 60 ng of

11674 Effect of Retinal PDEy on Transducin

FIG. 1. Effect of PDEy on PDE ac- tivity and GTPase activity. A, tryp- sin-activated PDE (8) was assayed for PDE activity with varying amounts of PDEy (50,51). The assay was for 10 min at 30 “C using 42 ng of PDE preparation per assay and 0-560 ng of Py prepara- tion. Based on densitometric scans (see “Experimental Procedures”) and esti- mated molecular weights, the PDEa was 0.002 nmol. At the 15,, point (-60 ng) the Py was estimated to be 0.006 nmol. Re- sults are means of triplicate samples. Percent inhibition indicates the activity compared to control PDE activity with- out added PDEy. B, GTPase activity was measured as described under “Experi- mental Procedures.” The assay con- tained 0.785 pg of Ta and 0-11.2 pg of PDEy per assay. The Ta constituted 6.8% of the total rod outer segment prep- aration which was added at 12.5 pgltube. Based on densitometric scans and esti- mated molecular weights, this is calcu- lated to be 0.021 nmol of Ta and 0.64 nmol of PDEy. Assays were for 20 min at 37 “C using 20 PM GTP and 1 pCi of [y-32P]GTP/tube in a final volume of 100 pl. Results are the mean of triplicate samples. Percent inhibition indicates the percent inhibition of GTPase activity of Ta compared to a control without added PDEy. Histone is type I1 from calf thy- mus (Sigma). Standard deviations (S.D.) were =+lo%.

PDE -Gamma (nanograms prote in)

B

2 4 I I 1 1 1 2

pg PDE - Gamma

PDEy. Based on densitometric scans of autoradiograms from Western blots (see above) 50% inhibition of PDE activity occurred at a PDEy:PDEa/P ratio of 3:l. This is close to the predicted ratio of 1:l and may be higher due to a 30% contamination of the PDE preparation by Ta.

Fig. 1B depicts the inhibition of Ta GTPase activity by the same PDEy preparation. In this undepleted rod outer segment preparation the IsO for PDEy inhibition of Ta GTPase activity was around 6 pg at a Ta:PDEy ratio of 1:30.

The actual Iso may be much lower since this preparation contained 6.8% Ta and considerable PDEaIP. The inhibition of GTPase activity by PDEy was not the result of nonspecific polycationic effects as histone did not inhibit GTPase activity at up to 200 mg (Fig. 1B).

Inhibition of the GTPase Activity of T a by PDEy Peptides- We have previously described the synthesis of peptides of bovine PDEy (18). These peptides were designated PDEy-1 through PDEy-6. Each of these peptides was tested for its

ability to inhibit PDE catalytic activity and Ta GTPase activity. The peptide, PDEy-3, which corresponds to bovine amino acid residues 31-45, was found to be the most effective peptide in inhibiting the GTPase activity of Ta (18). Fig. 2 depicts the dose-response curve for PDEy-3. The IsO, the quantity of peptide needed for 50% inhibition of GTPase activity, was 450 p ~ . This is about 100-fold higher than that for the native PDEy protein (Fig. 1). The inhibition of the GTPase activity of Ta by PDEy-3 was noncompetitive, caus- ing a decrease in the Vmax from 9 to 5 pM/min in the presence of 75 nmol of peptide, leaving the K , unchanged (18). This inhibition was not due to polycationic effects as polylysine had no inhibitory activity (Fig. 2).

Effect of PDEy-3 on the Release of GTPyS Bound to T a in the Absence of Rod Outer Segment Membranes-To further confirm the previously reported kinetic data (18), the effect of PDEy-3 on the release of GTPyS prebound to Ta in the absence or presence of GDP and in the absence of rod outer

Effect of Retinal PDEr on Transducin 11675

/ + P - Y - 3

n mole Peptide

FIG. 2. Effect of PDEy peptides on the GTPase activity of Ta. Undepleted rod outer segment membranes (4 pg) were preincu- bated 5 min with and without various quantities of the PDEy-3 peptide or polylysine. The reaction was initiated by the addition of GTP (GTP + 0.1 pCi of [y-32P]GTP) to a final concentration of 20 ~ L M in a volume of 100 p1 of GTPase buffer. After incubation at 37 "C for 20 min, released 32Pi was precipitated and quantitated as described under "Experimental Procedures." Results are the means from trip- licate samples. Standard deviations (S.D.) were ~ + 1 0 % . % inhibition indicates the percent inhibition of the GTPase activity of Toc com- pared to a control without added peptide.

TABLE I Effect of PDEy-3 on the release of P'SIGTPyS from Toc in the

absence of rod outer segment membranes Ta was labeled with [35S]GTPyS as described (24). The resulting

supernatant, which contains partially purified Ta-[35S]GTPyS, was incubated with various quantities of PDEy-3 for 10 min at 37 "C in a total volume of 100 pl of GTPase buffer. Bound nucleotide was determined as described under "Experimental Procedures." The re- sults are from triplicate samples * S.D. p values were calculated using Student's t test using 0 nmol of peptide as a reference. In this assay, 14 nmol = 140 p ~ .

PDEy-3 [36S]GTPyS bound

nmol nmol 0 1.84 + 0.15 I

14 1.96 * 0.05"

36 1.94 f 0.07"

72 2.11 * 0.05*

110 2.20 * 0.13'

140 2.02 * 0.02" 2.09 k 0.09"

0.1 > p > 0.05. ' 0.05 > p > 0.02.

segment membranes was determined (Table I). Ta was pre- labeled with [35S]GTPyS as described under "Experimental Procedures." The final purity of transducin in the resulting supernatant (partially purified TCX-[~~S]GTPYS) was greater than 85% as judged by SDS-PAGE (7.5% acrylamide) and Coomassie Blue staining. The partially purified TCX-[~~S] GTPyS was incubated in the presence of various quantities of the peptide, PDEy-3, and the amount of [35S]GTPyS remaining bound to Ta was determined using a filter binding assay. These results indicated that PDEy-3 did not decrease the amount of [35S]GTPyS prebound to Ta in the absence of rod outer segment membranes, nor did it compete with [35S] GTPyS in the absence of rod outer segment membranes. In a similar experiment, [3H]GDP was added to the reaction mixture at concentrations ranging from 2 to 100 FM in the

absence or presence of 100 nmol of PDEy-3 and incubated with the partially purified Ta-[35S]GTPyS. These results indicated that there was no exchange of [3H]GDP for [35S] GTPyS prebound to Ta in the presence of PDEy-3 and in the absence of rod outer segment membranes (data not shown).

Effect of PDEy-3 on Nucleotide Binding by Ta-Because the binding of GTPyS to Ta was not altered in the absence of rod outer segment membranes, the inhibition of the GTPase activity of Ta may have resulted from an inhibition of the initial GTP/GDP exchange on the rod outer segment membrane. To determine if this was the case, the extent of labeling of Ta with a nonhydrolyzable analog of GTP, GMPP(NH)P, was determined (Fig. 3). The addition of in- creasing quantities of PDEy-3 to undepleted rod outer seg- ment membranes resulted in a dose-response curve that was similar to the dose-response curve obtained for the inhibition of the GTPase activity of Ta by PDEy-3. The IS0 (the quantity of peptide needed for 50% inhibition of GMPP(NH)P binding by Ta) was about 550 PM, compared to the 150 of 450 FM for the inhibition of the GTPase activity of Ta. These results suggested that the peptide, PDEy-3, and possibly PDEy, inhibited the GTPase activity of Ta by altering the GTP/ GDP exchange on the rod outer segment membranes.

In a similar experiment, the effect of PDEy-3 on the

0 20 40 60 80 100 120 140 160 180 200

nmol PDEy - 3 FIG. 3. Effect of PDEy-3 on GMPP(NH)P Binding by Ta.

Undepleted rod outer segment membranes (50 pg) were preincubated 5 min with and without PDEy-3. The reaction was initiated by the addition of GMPP(NH)P substrate (GMPP(NH)P + 0.1 pCi of [3H] GMPP(NH)P) to a final concentration of 20 pM in a total volume of 100 p l of GMPP(NH)P-binding buffer. After incubation at 25 "C for 30 min, bound nucleotide was quantitated as described under "Ex- perimental Procedures." Results are the means from duplicate sam- ples. Percent inhibition indicates the percent inhibition of GMPP(NH)P binding by Toc compared to a control without added PDEy-3. The IS0 was about 55 nmol of PDEy3 or 550 p ~ .

11676 Effect of Retinal PDEy on Transducin

binding of [3H]GDP and [35S]GTPyS to Ta, both separately and together, was determined (Table 11). When [3H]GDP (1 pM final concentration) was incubated in the presence of undepleted rod outer segment membranes (previously dark- adapted, then exposed to room light for 5 min) and 200 nmol of PDEy-3, there was an increase in the amount of [3H]GDP bound to Ta from 0.54 to 0.72 pmol (33% increase). When [36S]GTPyS was substituted for [3H]GDP (1 p~ final concen- tration), the amount of [35S]GTPyS bound to Ta was about 10 times higher than the amount of [3H]GDP bound to Ta, because the affinity of GTPyS is much higher for Ta in the light than the affinity of GDP for Ta (35). In the presence of 200 nmol(2 mM) PDEy-3, the amount of [3SS]GTPyS bound to Ta was decreased from 8.6 to 6.6 pmol (22% decrease). In the presence of both nucleotides (1 PM final concentration each), no [3H]GDP binding to Ta was detected. In the pres- ence of 200 nmol of PDEy-3, the amount of [35S]GTPyS bound to Ta was decreased from 11 to 8.4 pmol (25% de- crease).

Effect of PDEy-3 on the Binding of Transducin to Rod Outer Segment Membranes-The inhibition of the exchange of GTP for GDP bound to Ta in the presence of the peptide, PDEy- 3, may be due to an alteration of the binding of Ta to rhodopsin (photoisomerized or nonphotoisomerized), TBy, or both rhodopsin and TPy on the surface of the rod outer segment membrane. To determine whether this was the case, the effect of PDEy-3 on membrane binding of transducin was determined.

We have used synthetic peptides to raise antisera to a wide variety of retinal rod outer segment proteins. These proteins include rhodopsin, Ta, TP and Ty (P- and y-subunits of transducin), S-antigen (48-kDa protein), interstitial retinol binding protein (IRBP), PDEa, and PDEy. These antisera were used to determine the purity and quantity of specific proteins in our crude transducin preparation and depleted rod outer segment membrane preparation using a solid-phase RIA.

To determine the effect of PDEy-3 on the association of Ta@y with rod outer segment membranes (interaction with rhodopsin), we have utilized a crude transducin preparation reconstituted with a depleted rod outer segment membrane preparation. The depleted membrane preparation was tested for the presence of specific rod outer segment proteins using antipeptide antisera and a solid-phase RIA. The depleted rod outer segment membranes tested positive for rhodopsin and negative for S-antigen, IRBP, PDEa, and PDEy. In this

TABLE I1 Effect of PDEy-3 o n PHIGDP and [36S]GTPyS binding by T a Dark-adapted, undepleted rod outer segment membranes (60 pg)

were incubated with and without 200 nmol (2 mM) PDEr-3 in the dark. Nucleotide was added to a final concentration of 1 p M in a total volume of 100 p1 of GMPP(NH)P-binding buffer. The reaction was initiated by illumination of the rod outer segment membranes and incubating for 5 min at 37 ‘C. Bound nucleotide was determined as described under “Experimental Procedures.” Results are from tripli- cate samples f S.D. p values were determined by the standard Student’s t test and were less than 0.05 for all samples & peptide. ND = not detected. Values in parentheses are percent inhibition of nucleotide binding in the presence of peptide. ~

Sample PDEr-3 [3H]GDP [Y3]GTP?S pmol bound

ROS + [3H]GDP ROS + [3H]GDP + 0.72 f 0.03 ROS + [36S]GTPrS - ROS + [36S]GTPyS

8.6 f 0.6 + 6.7 f 0.2

ROS + both nucleotides - ND 11 f 1 ROS + both nucleotides + ND 8.4 f 1.3

- 0.54 k 0.07

(22)

(25)

depleted rod outer segment membrane preparation, rhodopsin constituted 85% of the total protein as judged by SDS-PAGE. The crude transducin preparation was analyzed in a similar manner. The preparation tested negative for rhodopsin, IRBP, and S-antigen, and tested positive for Ta and TPy. PDEaP2 was also present in the crude transducin preparation but in smaller amounts than transducin. The crude transducin preparation was 80% Tap?, and 10% PDEaPy, as judged by

The effect of PDEy-3 on the association of transducin with rod outer segment membranes was determined by incubating the crude transducin preparation with GTP, in the presence of light, with and without PDEy-3, and in the presence or absence of the depleted rod outer segment membranes. In a typical assay, 18 pg of rhodopsin and 8 pg of the crude transducin preparation were incubated for 10 min in the light with 40 p~ GTP in the presence or absence of 51 nmol (510 p ~ ) PDEy-3. The ratio of rhodopsin to transducin was about 2:l based on total protein concentration of the rod outer segment membrane preparation and the crude transducin preparation (the actual ratio may be higher since the crude transducin preparation was only 80% pure). The quantity of peptide added was at least 10-fold greater than the quantity of transducin and rhodopsin added to the assay mixture. This may be necessary because only a small amount of the peptide may be in the correct conformation in solution to bind to transducin. Circular dichroism measurement of the peptide PDEy-3 has shown that the peptide appears to be in a random coil conformation in solution (Fig. 4). This is different than the mostly a-helical conformation of PDEy-3 predicted by the Chou and Fasman (19) method for the intact protein.

After incubation for 10 min, the rod outer segment mem- branes were sedimented by centrifugation, and the resulting supernatant was drawn off with a Pasteur pipette and tested for specific proteins using peptide antisera and the solid- phase RIA.

Fig. 5 shows the results of an experiment in which Ta was quantitated by a solid-phase RIA using antisera to a peptide called Ta-C. This peptide corresponds to the last 15 amino acids of the C terminus of bovine Ta. All of these experiments were performed in room light and with 40 pM GTP. As the results indicate, centrifugation of the reaction mixture does not remove transducin from the crude tranducin preparation

SDS-PAGE.

- -

190 260 2io 210 240 : 0

Wavelength A (nm)

FIG. 4. Circular dichroism spectrum for PDEy-3. Purified peptide was dissolved in 0.01 M sodium phosphate, 2 mM MgCl,, pH 7.0, to a final concentration of 0.5 mg/ml. The peptide was scanned from 240 to 195 nm at a rate of 10 nm/min. with a time constant of 4, in a 0.5-mm quartz cell a t room temperature. The resulting circular dichroism spectrum was compared to standards determined by Green- field and Fasman (39).

Effect of Retinal PDEy on Transducin 11677

FIG. 5. Effect of PDEy-3 on the interaction of Tcu with rod outer segment membranes. Transducin (T), depleted rod outer segments ( R O S ) , or both transducin and depleted rod outer segments were incubated in the absence or presence of 51 nmol (510 KM) of PDEy-3 ( P y 3 ) at room temperature for 10 min in a total volume of 100 p1 of GTPase buffer. The samples were centrifuged, and the resulting supernatants were tested for the quantity of Ta using a solid-phase RIA and antipeptide antisera for Ga-C, a peptide corre- sponding to the last 15 amino acids of the C terminus of bovine Ta. Results are from six samples (mean & S.D.). The binding of Ta to equal micrograms of lens membranes was used as a negative control This value has been subtracted.

in the absence of membranes (T), nor does the depleted rod outer segment membrane preparation contribute any trans- ducin to the supernatant after centrifugation when incubated in the absence of the crude transducin preparation (ROS). When the crude transducin preparation and the depleted rod outer segment preparation are incubated together, some of the Ta remains bound to the depleted rod outer segments and is, therefore, removed from the supernatant when the mem- branes are centrifuged (T+ROS). When PDEy-3 is added to the crude transducin preparation and the depleted rod outer segment membranes, essentially all of the Ta is removed from the supernatant after centrifugation of the membranes (T+ROS+Py3). The presence of PDEy-3 appears to enhance the binding of Ta to the rod outer segment membrane. This is not due to PDEy-3 causing the sedimentation of Ta in the absence of rod outer segment membranes or interference of the interaction of Ta-C with Ta because the quantity of Ta in the supernatant is unchanged in the presence of PDEy-3 after centrifugation (T+Py3). Also, the amount of Ta is not increased when depleted membranes are incubated in the presence of PDEy-3 (ROS+Py).

In a similar experiment, the quantity of TP was determined by using antisera to a peptide called TP-C. This peptide corresponds to the last 15 amino acids of the C terminus of TB. These results were similar to those for Ta, except that more TB was bound to the depleted rod outer segment mem- branes in the absence of PDEy-3, leading to a smaller decrease in the amount of TP in the supernatant in the presence of PDEy-3. Also, an analog of PDEy-3, found not to inhibit the GTPase activity of transducin (Analog 1: see Table I11 and Fig. 8A) , was ineffective in inducing the binding of Ta to the rod outer segment membrane (data not shown). Polylysine, when added at 100 mM, likewise, had no effect.

Iz5I-PDEy-3+Tyr Binding to Rod Outer Segment Mem- branes-To determine whether PDEy-3 bound specifically to undepleted rod outer segment membranes, a synthetic analog of PDEy-3, PDEy-3+Tyr, was used. This analog has a tyro- sine on the N terminus to allow for radioiodination by the

TABLE 111 PDEy-3 analogs

PDEy-3 corresponds to bovine PDEy amino acid residues 31-45. Starred amino acids are substitutions from the original PDEy-3 sequence used to make the indicated analog. PDEy-a-helix corre- sponds to the region extending from amino acid residues 29-41 predicted to be a-helical by the methods of Chou and Fasman (19).

PDEy-3 H3N-K-Q-R-Q-T-R-Q-F-K-S-K-P-P-K-K-COOH

Analog 1 H,N-K-Q-R-P-T-R-Q-F-K-S-K-P-P-K-K-COOH

Analog 2 H.N-K-Q-R-E-T-R-Q-F-K-S-K-P-P-K-K-COOH

Analog3 H,N-K-Q-Q-Q-T-R-Q-F-K-S-K-P-P-K-K-COOH

Analog4 H,N-K-Q-R-Q-V-R-Q-F-K-S-K-P-P-K-K-COOH

Analog5 H,N-K-Q-R-Q-T-E-Q-F-K-S-K-P-P-K-K-COOH

Analog6 H,N-K-Q-R-Q-T-R-Q-L-K-S-K-P-P-K-K-COOH

Analog 7 HSN-K-Q-R-Q-T-K-Q-F-K-S-K-P-P-K-K-COOH

Analog8 H,N-K-Q-R-Q-T-R-Q-F-K-V-K-P-P-K-K-COOH

Analog9 H3N-K-Q-R-N-T-R-Q-F-K-S-K-P-P-K-K-COOH

Analog 10 H,N-K-Q-R-Q-T-R-Q-E-K-S-K-P-P-K-K-COOH

Analog 11 H3N-K-Q-R-Q-T-R-E-F-K-S-K-P-P-K-K-COOH

Analog 12 H3N-K-Q-R-Q-T-R-Q-F-K-S-K-P-Q-K-K-COOH

Analog 13 H3N-K-Q-R-Q-T-R-P-F-K-S-K-P-P-K-K-COOH PDEy-a-helix H,N-K-F-K-Q-R-Q-T-R-Q-F-K-S-K-COOH

*

*

e

*

*

*

*

*

chloramine-T method to make the labeled peptide '251-PDEy- 3+Tyr as described above. PDEy-s+Tyr inhibited the GTPase activity of Ta only slightly less than PDEy-3 (35% inhibition at 510 I.LM of PDEy-3+Tyr compared to 50% inhi- bition at 450 I.LM of PDEy-3). When PDEy-3+Tyr was iodi- nated using unlabeled iodine, the peptide still inhibited the GTPase activity of Ta by 35%. The buffers used in the iodination of PDEy-3+Tyr did not inhibit the GTPase activ- ity of transducin alone (data not shown).

Fig. 6 illustrates the results of the binding of '251-PDEy- 3+Tyr to undepleted rod outer segment membranes, unde- pleted lens membranes, and to the soluble protein BSA. BSA was added above the level that was already present in the buffer. The results indicated that at least twice as much PDEy-3+Tyr bound to the undepleted rod outer segment membranes than to either lens membranes or BSA. The large amount of binding exhibited by the lens membrane may be due to nonspecific binding by either residual free lZ5I or lZ5I- PDEy-3+Tyr. The binding of lZ5I-PDEy-3+Tyr to the unde- pleted rod outer segment membranes was inhibited by unla- beled PDEy-3 and PDEy-3+Tyr but not by IRBP-1, a peptide corresponding to bovine IRBP residues 1-15 (Fig. 7).

Effect of PDEy Analogs on the GTPase Activity of Ta-In our previously published report (18), we analyzed the amino acid sequence of PDEy for predicted secondary structures based on the statistical method of Chou and Fasman (19) and determined the hydrophobicity plot based on the method of Kyte and Doolittle (20). These analyses indicated a region extending from amino acid residues 29 to 41, which was predicted to be a-helical and hydrophilic in the intact PDEy subunit. This region was flanked by 4 prolines and contained a large number of positively charged amino acids. The pre- dicted structure of this region placed all of the positive charges on one surface of the a-helix, while hydrophobic residues were

11678 Effect of Retinal PDEr on Transducin

280-

240-

200-

L >r

+ - 160-

0 I

W r 120-

I 80-

N - 40-

mg protein FIG. 6 . ‘Z61~PDEy-3 binding to membrane homogenates.

Peptides were radioiodinated as described under “Experimental Pro- cedures.” The radioiodinated peptide, lZ6I-PDEy-3+Tyr (1.2 nmol; 12 p ~ ) was incubated with various quantities of undepleted rod outer segment membranes (O), lens membranes (0), or BSA (X) in 100 pl of GTPase buffer + 1 mg/ml BSA. After incubation at 37 “C for 10 min, bound peptide was determined as described under “Experimental Procedures.” Results are the means from duplicate samples.

z 0 0

E c?

I

d I 0 l b io &I 40

nmol peptide

FIG. 7. Competition of ““I-PDEy-3 binding by unlabeled peptides. Undepleted rod outer segments were incubated with var- ious quantities of unlabeled PDEy-3 (0), PDEr-B+Tyr (X), and IRBP-l(O). The radioiodinatedpeptide, ‘Z61-PDEr-3+Tyr (1.2 nmol,

volume of 100 p1 in GTPase buffer + 1 mg/ml BSA. After incubation 12 p ~ ) was added to the rod outer segment membranes to a total

at 37 “C for 10 min, bound peptide was quantitated as described under “Experimental Procedures.” Results are the means from duplicate samples.

located on the other surface. Most of this predicted a-helical region was located in the peptide, PDEy-3, which extends from amino acid residues 31 to 45. We used the predicted

structure to determine which amino acids could be substituted in PDEy-3 for the synthesis of PDEy-3 analogs. These ana- logs were used to determine functionally important amino acids needed for PDEy-3 binding to Ta and for the inhibition of the GTPase activity to Ta.

Table I11 lists the 13 PDEy-3 analogs which were synthe- sized, and a peptide corresponding to the predicted a-helical region called PDEy-a-helix (bovine amino acid residues 29- 41). In these analogs, amino acids were substituted to affect the structure of the peptide, to alter the positive charges on the hydrophilic surface or to replace hydrophobic amino acids with hydrophilic or charged amino acids. Amino acids on the hydrophobic surface of the predicted a-helix which were sub- stituted included glutamine 34 and phenylalanine 36. Posi- tively charged amino acids on the hydrophilic surface which were substituted included arginine 33 and arginine 36. Other amino acids substituted, which may be important in hydrogen bonding but contain no positive charge, included glutamine 37, threonine 35, and serine 40. Substitutions were made so as to affect the secondary prediction plot of the intact protein as little as possible (except for Analogs 1, 12, and 13), and, in some cases, to affect the hydrophobicity plot as little as possible (Analogs 1, 3, 6, 7, 9, 12, and 13). The peptides were then tested for their ability to inhibit the GTPase activity of Ta (Fig. 8, A and B ) .

A

70 ‘1

m o l peptide

1 B

‘0

nmol peptide

FIG. 8. Effect of PDEy-3 analogs on the GTP- activity of transducin. Assays were performed as described in Fig. 2. Results are the means from triplicate samples. Standard deviation values were <+IO%. Peptides tested were: A, 0, Analog 1; 0, Analog 2; X, Analog 3; W, Analog 4; 0, Analog 5; A, Analog 6; and A, Analog 7. B, 0, Analog 8; 0, Analog 9; X, Analog 10; W, Analog 11; 0, Analog 12; A, Analog 13; A, PDEy-a-helix. PDEr-3 (V) was included in each graph for reference.

Effect of Retinal PDEr on Transducin 11679

Analogs 1, 12, and 13 were made to attempt to disrupt the structure of the peptide, PDEy-3. In analogs 1 and 13, two different glutamines were substituted with prolines, one on the hydrophobic surface and one on the hydrophilic side, to disrupt any a-helical structure the peptide may obtain upon binding to transducin. These substitutions were made such that the hydrophobicity was not substantially altered. The third peptide, Analog 12, had a proline replaced by a glutamine at the site where there are normally two flanking prolines. This would determine the importance of the flanking prolines in the maintenance of the structure of PDEy-3, when it is bound to transducin. Circular dichroism measurements of these analogs suggested that they were all in a random coil conformation in solution (data not shown). The circular di- chroism spectra for all analogs were identical to the parent peptide PDEy-3 (see Fig. 4).

When glutamines were replaced with prolines in the region predicted to be a-helical (Analogs 1 and 13), the inhibition of the GTPase activity of Ta by the peptide analogs was reduced. This suggested that the structure of the peptide may be disrupted by the substitution of the glutamines with prolines. If the structure of this region is indeed a-helical in the intact PDEy peptide or when the peptide, PDEy-3 is bound to Ta, then the addition of the prolines could make it more difficult for the peptide to obtain this structure upon binding to transducin. Addition of a glutamine in place of one of the prolines flanking the predicted a-helical region did not de- crease the inhibitory effectiveness of the peptide, PDEy-3, as much as the substitution of amino acids in the predicted a- helical region. This suggests that the structure of this region in PDEy may depend, in part, on the amino acids in the flanking regions.

Since one side of the predicted a-helix contains a large number of positively charged amino acids (arginines and lysines), substitution of these amino acids would determine the functional important of these amino acids in the inhibition of the GTPase activity of Ta by the peptide, PDEy-3. When either of the arginines in the predicts a-helical region were substituted with glutamine and glutamic acid (Analogs 3 and 5, respectively), the peptide's ability to inhibit the GTPase activity of Ta was greatly reduced (Iso > 2 and 2.5 mM, respectively) when compared to the peptide, PDEy-3 (160 = 450 pM). When a glutamine adjacent to one of these arginines was replaced by a negatively charged glutamic acid (Analog 11) to neutralize the charge on the adjacent arginine, the decrease in the inhibitory effectiveness of the peptide was not as great as when the arginine was totally replaced (Analogs 3 and 5). When one of these arginines was replaced by another positively charged amino acid (Analog 7), lysine, the ability of the peptide to inhibit the GTPase activity of Ta was not decreased. These results indicate that the positively charged amino acids may be necessary for the binding of PDEy-3 to tranducin and for the inhibition of the GTPase activity to Ta. In contrast, peptides with substitutions made on the hydrophobic surface of the predicted a-helical region (Analogs 2, 6, 9, and 10) were still able to inhibit the GTPase activity of Ta as well as PDEy-3. When a serine or a threonine was substituted with a valine (Analogs 4 and 8, respectively), the inhibition of the GTPase activity of transducin was reduced, suggesting that there may be other hydrophilic interactions in the binding of PDEy-3 to transducin and for the inhibition of the GTPase activity of Ta.

An analog corresponding to the predicted a-helical region was also tested for its ability to inhibit the GTPase activity of Ta (Fig. 8B). This peptide inhibited the GTPase activity of Ta as effectively as PDEy-3. This suggests that the pre-

dicted a-helical region alone is able to interact with transducin without the flanking prolines. Although the flanking prolines may not be necessary for the interaction of the predicted a- helical region with transducin, the presence of these prolines may play an important role in maintaining the secondary structure of this region in the intact PDEy subunit.

DISCUSSION

The results of the above experiments (Fig. 1) and those previously reported (18) suggest that PDEy might regulate the GTPase activity of Ta and, ultimately, the shut-off mech- anism of the visual transduction system. Others have sug- gested that PDEy may interact directly with Ta (12, 13, 30, 31).

The peptide, PDEy-3, which corresponds to bovine PDEy amino acid residues 31-45, contains part of a region of PDEy which effectively inhibits the GTPase activity of Ta (Fig. 2). Lipkin et at. (40) have reconfirmed our earlier results (18) that the region for the interaction of PDEy with Ta resides in amino acid residues 25-47. We have shown that this inhibitory region may actually extend between amino acid residues 29-41, as shown by the more specific results using the synthetic peptide PDEy-a-helix (Table 111) in the GTPase assay (Fig. 8B). This region was predicted to have an a-helical structure containing a large number of positively charged amino acids and amino acids which may participate in hydro- gen bonding on the hydrophilic surface of the predicted a- helix. These amino acids may be important for the interaction of PDEy-3 with transducin as shown by the analog studies described in Fig. 8, A and B. The inhibition of the GTPase activity of transducin is noncompetitive (18) and may result from an alteration of the interaction between transducin a and the rod outer segment membrane, interfering with the GTP/GDP exchange on Ta. This is suggested by the results of the previously reported kinetic experiments (18), by the

PI

FIG. 9. GTP hydrolytic cycle. The model of the GTP hydrolytic cycle was based on several studies (35, 42-45). Numbers indicate the mechanisms described in the text. Symbols used in the model are: R*, photoexcited rhodopsin; T, transducin-a@y (Ta@y) complex; R*-T'- GDP, closed conformation of Ta-GDP bound to rhodopsin as an R*- TD-yTd-GDP complex; R*-To-GDP, R*-T", and R*-T"-GTP, open conformations of Ta with GDP, no nucleotide, and GTP-bound, respectively, cGMP-phosphodiesterase. Numbers indicate the mech- anisms discussed under "Discussion."

11680 Effect of Retinal PDEy on Transducin

inhibition of guanine nucleotide binding by Ta in the presence of membranes (Fig. 3 and Table 11), and by the enhanced binding of transducin to the rod outer segment membrane as shown by the reconstitution experiments (Fig. 5).

The steps leading to the exchange of GTP for GDP pre- bound to Ta and the subsequent release of Ta-GTP exist in the equilibrium shown in Fig. 9 (35, 42-46). The equilibrium is shifted counterclockwise by the binding of GTP to Ta and by the dissociation of Ta-GTP from photoexcited rhodopsin (R*) (Keq > 100) (42, 43). The affinity of transducin for R* depends on whether Ta has GTP, GDP, or no nucleotide bound (35,36,41). Hingorani and Ho (32) have also proposed a molecular model for the activation of Ta which has also been incorporated into the model depicted in Fig. 9. In this model, Ta exists in a “closed” state (Te) when bound to nonphotoisomerized rhodopsin (R*-T0yTaC-GDP = R*-Te- GDP), in which the GTP/GDP binding site is nonaccessible to guanine nucleotides. When rhodopsin is photoisomerized, the GTP/GDP binding site “opens” (TO) (R*-T”-GDP and R*-T”-: see Fig. 9) and allows for the exchange of GTP for GDP bound to Ta (R*-To-GTP). This two-state model has also been proposed by others (35,44). After guanine nucleotide exchange, Ta-GTP is released from T@y and R* and can activate PDE. The forward reactions of these equilibria are driven by the hydrolysis of GTP by Ta (41).

In the undepleted rod outer segment preparations used in the GTPase assays, most of the transducin is in the R*-To- GDP or R*-To state because the preparations were exposed to light prior to their use in the GTPase assays. Preparations of rod outer segment membranes will hydrolyze most of the GTP to GDP and will remain bound to R* as long as exoge- nous GTP is not present (26, 34). This factor was taken into account when developing a possible mechanism for the inhi- bition of the GTPase activity of Ta by PDEy.

There are four possible mechanisms by which the peptide, PDEy-3, and possibly PDEy, may inhibit the GTPase activity of Ta. These are: 1) decreasing the intrinsic catalytic rate of GTP hydrolysis by Ta through a mechanism not affecting the GTP/GDP exchange, 2) preventing the association of TpyTa-GDP (T-GDP) with R* by enhancing the binding of Ta-GDP with PDEa0, 3) preventing the association of T - GDP with R* by the enhanced binding of T-GDP to light and GTP insensitive sites on the rod outer segment membrane, and 4) increased binding of T-GDP to R* accompanied by an interference of GTP/GDP exchange (37,38).

Based on the results reported herein, the inhibition most likely occurs by the increased affinity of Ta for TPy which could result in an increased affinity of transducin for R*. One function of TPy may be to facilitate the reassociation of Ta- GDP with R* (33, 45). It has been suggested that GPy from the adenylate cyclase system may be indirectly involved in the inhibition of adenylate cyclase activation by Ga, by in- creasing the binding of Gas to 0-adrenergic receptor and preventing it from associating with and activating adenylate cyclase (47). The same mechanism could regulate the activa- tion of PDE by Ta-GTP. In the molecular model of Hingorani and Ho (32) the binding of GTP to the guanine nucleotide binding domain of Ta moves two a-helices into the receptor binding domain which causes the simultaneous release of T a - GTP from R* and T@y. The role of PDEr may be to increase the affinity of Ta-GDP for TPy which could lock the receptor binding domain of Ta into the R*/TPy bound conformation preventing the movement of the two a-helices from the gua- nine nucleotide binding domain into the receptor binding domain. The overall effect of PDEy on Ta may be a slight

modulation of the cascade, depending on the localized concen- trations of each protein within the outer segment.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15. 16. 17.

18.

19.

20. 21.

22.

23.

24.

25. 26. 27.

28. 29. 30.

31.

32. 33. 34.

35.

36. 37.

38.

39.

40.

41.

REFERENCES Bitensky, M. W., Gorman, R. E., and Miller, W. H. (1971) Proc.

Cote, R. H., Bierbaum, M. S., Nicol, G. D., and Bownds, M. D.

Nicol, G. G., and Miller, W. H. (1978) Proc. Natl. Acad. Sci. U. S. A. 75 , 5217-5220

Fesenko, E. E., Kolesnikov, S. S., and Lyvbarsky, A. L. (1985) Nature 313 , 310-313

Cook, N. J., Zeilinger, C., Koch, K. W., and Kaup, U. B. (1986) J. Biol. Chem. 261,17033-17039

Cook, N., Hanke, W., and Kaup, B. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,585-589

Baehr, W., Devlin, M. J., and Applebury, M. L. (1979) J. Biol. Chem. 254,11669-11677

Hurley, J. B., and Stryer, L. (1982) J. Biol. Chem. 267, 11094- 11099

Ovchinnikov, Y. A., Lipkin, V. M., Kumarev, V. P., Gubanov, V. V., Khramstov, N. V., Akhmedov, N. B., Zagranichny, V. E., and Muradov, K. G. (1986) FEBS. Lett. 204,288-292

Fung, B. K-K., Hurley, J. B., and Stryer, L. (1981) Proc. Natl. Acad. Sci. U. S. A. 7 8 , 152-156

Sitaramayya, A., Harkness, J., Parkes, J. H., Gonzalez-Olivia, C., and Liebman, P. A. (1986) Biochemistry 25,651-656

Whalen, M., and Bitensky, M. (1989) Biochem. J. 2 6 9 , 13-19 Griswold-Prenner, I., Young, J. H., Yamane, H. K., Fung, B. K-

K. (1988) Znuest. Ophthalmol. & Visual Sci. 2 9 , (suppl.) 218 Hodges, R. S., and Merrifield, R. B. (1975) Anal. Biochern. 6 5 ,

Gormann, J. J. (1984) Anal. Biochem. 136,397-406 Merrifield, R. B. (1964) Biochemistry 3, 1385-1390 Lookhart, G. L., Jones, B. L., Cooper, D. B., and Hall, S. B.

Morrison, D. F., Rider, M. A., and Takemoto, D. J. (1987) FEBS

Chou, P. Y., and Fasman, G. D. (1978) Annu. Rev. Biochem. 4 7 ,

Kyte, J., and Doolittle, R. F. (1982) J. Mol. Bwl. 157 , 105-132 Papermaster, D. S., and Dreyer, W. J. (1974) Biochemistgv 1 3 ,

Hamm, H. E., and Bownds, M. D. (1984) J. Gen. Physiol. 8 4 ,

Takemoto, D. J., Haley, B. E., Hansen, J., Pinkett, O., and Takemoto, L. J. (1981) Biochem. Biophys. Res. Commun. 102,

Stryer, L., Hurley, J. B., and Fung, B. K-K. (1983) Methods

Laemmli, U. K. (1970) Nature 227,680-685 Kuhn, H. (1980) Nature 283,587-589 Russell, P., Robison, W. G., and Kinoshita, J. H. (1981) Exp. Eye

Suter, M. (1982) J. Zmmunol. Methods 53. 103-108 Bradford, M. M. (1976) Anal. Biochem. 7 2 , 248-254 Hingorani, V. N., Tobias, D. T., Henderson, J. T., and Ho, Y-K.

Deterre, P., Bigay, J., Pfister, C., Kuhn, H., and Chabre, M.

Hingorani, V. N., and Ho, Y-K., (1987) FEBS Lett. 2 2 0 , 15-22 Fung, B. K-K. (1983) J. Biol. Chem. 258,10495-10502 Fung, B. K-K., and Stryer, L. (1980) Proc. Natl. Acad. Sci.

Bennett, N., and Dupont, Y. (1985) J. Bwl. Chem. 260, 4156-

Rybin, V. 0. (1986) Biokhimiya 51,1035-1041 Miki, N., Baraban, J. M., Keirns, J. J., Boyce, J. J., and Bitensky,

Kuhn, H., and Hargrave, P. A. (1981) Biochemistry 2 0 , 2410-

Greenfield, N., and Fasman, G. D. (1969) Biochemistry 8,4108-

Lipkin, V. M., Dumler, I. L., Muradov, K. G., Artemyev, N. O.,

Stryer, L. (1985) Biopolymers 24,29-47

Natl. Acad. Sci. U. S. A. 68,561-565

(1984) J. Bid. Chem. 259,9635-9641

241-272

(1982) J. Biochem. Biophys. Method 7, 15-23

Lett. 222,266-270

251-276

2438-2444

265-280

341-347

Enzymol. 9 6 , 617-627

Res. 32,511-516

(1988) J. Biol. Chem. 263,6916-6926

(1986) Proteins Struct. Funct. Genet. 1, 188-193

U. S. A. 77,2500-2504

4168

M. W. (1985) J. Bid. Chem. 250,6320-6327

2417

4116

and Etingof, R. N. (1988) FEBS Lett. 234,287-290

Effect of Retinal PDE-y on Transduein 11681

42. Yamanaka, G., and Stryer, L. (1984) Invest. Ophthalmol. & Visual R. L., Birnbaumer, L., Spiegel, A. M., Caron, M. G., and Sci. 26, (suppl.) 157 Lef'kowitz, R. J. (1986) J. Biol. Chem. 261,9514-9520

43. Stryer, L. (1986) Annu. Reu. Neurosci. 9,87-119 48. McMurray, M. M., Hansen, J. S., Haley, B. E., Takemoto, D. J., 44. Kohl, B., and Hofmann, K. P. (1987) Biophys. J. 52, 271-277 and Takemoto, L. J. (1985) Biochem. J . 226,227-232

45. Yamazaki, A., Tatsumi, M., Torney, D. C., and Bitensky, M. W. 49. Spiegel, M. R. (1961) Theoryand Problems of Statistics, McGraw

(1987) J. Biol. Chem. 262,9316-9323 Hill, Inc., New York

46. Shichi, H. (1979) Biochemistry of Vision, Academic Press, New 50. Thompson, W. J., and Appleman, M. M. (1971) Biochemistry 10,

311-316 York 51. Wensel, T. G., and Stryer, L. (1986) Proteins Strut . Func. Genet.

47. Cerione, R. A,, Staniszewski, C., Gierschik, P., Codina, J., Somers, 1,90-99